Pod fails with error "java.net.NoRouteToHostException: No route to host (Host unreachable)" in Openshift Container Platform 3

Issue

The error below happens irregularly and affects all pods on that node.
From journal logs

atomic-openshift-node[2309]: I1029 12:49:29.554909    2309 prober.go:111] Readiness probe for "jenkins-12-5qdx6_cicd :jenkins" failed (failure): Get http://10.10.0.0:8080/login: dial tcp 10.12.0.0:8080: getsockopt: connection refused                                      
atomic-openshift-node[2309]: I1029 12:49:41.047119    2309 prober.go:111] Readiness probe for "jenkins-12-5qdx6_cicd:jenkins" failed (failure): HTTP probe failed with statuscode: 503                                                                                            │error while getting metrics summary from Kubelet node.example.com (10.10.0.0:10250): Get https://10.10.0.0:10250/stats/summary/: dial tcp 10.10.0.0:10250: getsockopt: no route to host

From pod

java.net.NoRouteToHostException: No route to host (Host unreachable)